Breaking T cell tolerance is arguably a critical step in the pathogenesis of SLE. Activated T cells provide help to autoantibody-secreting B cells and also infiltrate target organs. Therefore it is important to identify how, where and when such T cells are activated and how this is maintained. In fact, the latter is critical because any therapy will have to interrupt ongoing disease. Lupus-prone mice lacking B cells from birth have markedly reduced disease and much less T cell activation. This suggested that B cells are upstream in T cell activation and also that they would be good therapeutic targets. However, in lupus mice B cell depletion was difficult to achieve using standard anti-CD20 mAb and effects, particularly on T cell activation, were not striking. Similarly anti-CD20 failed to show clinical effect in SLE patients, though anti-BLyS did. These findings raise the question of what the role of B cells really was in ongoing disease, and whether, if efficient depletion were possible, would T cell activation and disease be reduced. They also suggested that other cell types such as DCs could be as or more important both for initiation and propagation. In this proposal we will use genetic approaches to target key cells and molecules-including deletion during disease rather than from birth-in order to unravel the complex interplays of disease mechanisms in vivo. To this end, we recently deleted cDC and pDC in lupus- prone MRL.Faslpr mice from birth. Surprisingly, we found little effect on T cell activation in secondary lymphoid tissue (SLT), but there were striking reductions in nephritis and skin disease, decreased proteinuria and longer survival. These data suggested that DCs might be required to activate T cells in target tissues-a novel role rather than to initiate activation, as might have been assumed. We propose a new model in DCs have non- redundant roles in tissue infiltration in lupus but that B cells are the most upstream APC in SLT. To address this model, we will first deplete both B cells and DCs during ongoing disease, rather than from birth. This will validate DCs as therapeutic targets and help resolve the controversial role of B cells. We will also distinguish the roles of cDCs from pDCs, as whether pDCs are important for SLE has not been tested experimentally. Second, to understand how B cells and DCs might be activating T cells, we will block APC function by deleting MHCII from each. Further, we will investigate the role of ICOS signals, which are linked to lupus in a number of ways, including emerging data that ICOS specifies T follicular helper (TFH) and T extrafollicular helper differentiation, cells that help autoreactive B cells. We also have exciting preliminary data that ICOSL on DCs specifically is required to promote infiltrating T cells in kidney, and that such cells express Bcl6 and thus may be re- lated to TFH. To further investigate the roles of ICOS signals, we will specifically delete ICOSL on B cells and on DCs. Together these experiments-organized around Aim 1 focusing on B cells and Aim 2 focusing on DCs-will test our model; determine if B cells/DCs or both are important as APC; validate B cells, cDCs, and pDCs as therapeutic targets; and elucidate the role of ICOS signals in the development of pathogenic T cells.